SQL将具有多列的表转换为两个表之间具有FK

Question

I need to convert my table with several fields to two other tables where one of the new tables has a row for each field in the first table in Microsoft SQL Server.

Table1(Table1Id, Field1, Field2, Field3)

for each row in Table1 create

Table2a(Table2aId)
Table2b(Table2bId, Table2aId, Field1)
Table2b(Table2bId, Table2aId, Field2)
Table2b(Table2bId, Table2aId, Field3)

Details
I currently have the table

Table1

[dbo].[CommunityAssetTemplates]  
      ,[CommunityId]  
      ,[CommunityAssetTemplateId]  
      ,[BaseHouseSpecsAssetId]  
      ,[CommunityLogoAssetId]  
      ,[CommunityMarketingMapAssetId]  
      ,[CommunityPhotoAssetId]  
      ,[CommunityVideoDraftAssetId]  
      ,[CommunityVideoAssetId]  

This was mostly a quick way to fulfill a business need before we fully implemented the new feature where users can define multiple templates with different assets in them, so I made two new tables one is just to relate the second table to a Community

Table2a

[dbo].[CommunityAssetDataTemplates]  
,[CommunityAssetDataTemplateId]    
,[CommunityAssetTemplateTypeId]  
,[CommunityId]  

Table2b

[dbo].[CommunityAssetTemplateFiles]  
,[CommunityAssetTemplateFileId]  
,[CommunityAssetDataTemplateId]  
,[CommunityAssetId]  

These two tables map together like so, each Table1 row creates 1 Table2a row and 6 Table2b rows

Table2a

[CommunityAssetDataTemplateId] Auto Increments  
[CommunityAssetTemplateTypeId] = 1  
[CommunityId]  = Table1.CommunityId  

Table2b - 1

[CommunityAssetTemplateFileId]  Auto increments  
    ,[CommunityAssetDataTemplateId] = Table2a.[CommunityAssetDataTemplateId]  
    ,[CommunityAssetId] = Table1.[BaseHouseSpecsAssetId] (THIS CHANGES)  

Table2b - 2

[CommunityAssetTemplateFileId]  Auto increments  
    ,[CommunityAssetDataTemplateId] = Table2a.[CommunityAssetDataTemplateId]  
    ,[CommunityAssetId] = Table1.[CommunityLogoAssetId] (THIS CHANGES)  

continues for the remaining 4 'AssetId's fields of Table1

Answer 1

Here is one way to accomplish this using CROSS APPLY to separate Field1 , Field2 , and Field3 columns into rows:

insert into Table2A (Table2Id)
select Table1Id from Table1

insert into Table2B(Table2Id, Field4)
select Table1Id, Field
from Table1
cross apply (values(Field1), (Field2), (Field3)) as ColumnsAsRows(Field)

Here is a sample:

declare @t1 table (Table1Id int identity(1,1), Field1 int, Field2 int, Field3 int)
declare @t2 table (Table2Id int primary key clustered)
declare @t3 table (Table3Id int identity(1,1) primary key clustered, Table2Id int, Field4 int)

insert into @t1 (Field1, Field2, Field3)
values (1, 2, 3), (4, 5, 6), (7, 8, 9)

select * from @t1

insert into @t2 (Table2Id)
select Table1Id from @t1

insert into @t3 (Table2Id, Field4)
select Table1Id, Field
from @t1
cross apply (values(Field1), (Field2), (Field3)) as ColumnsAsRows(Field)

select * from @t2
select * from @t3